Rural–Urban Differences in the Decline of Adolescent Cigarette Smoking

Abstract

Objectives. To examine change over time in cigarette smoking among rural and urban adolescents and to test whether rates of change differ by rural versus urban residence.

Methods. We used the 2008 through 2010 and 2014 through 2016 US National Survey of Drug Use and Health to estimate prevalence and adjusted odds of current cigarette smoking among rural and urban adolescents aged 12 to 17 years in each period. To test for rural–urban differences in the change between periods, we included an interaction between residence and time.

Results. Between 2008 to 2010 and 2014 to 2016, cigarette smoking rates declined for rural and urban adolescents; however, rural reductions lagged behind urban reductions. Controlling for socioeconomic characteristics, rural versus urban odds of cigarette smoking did not differ in 2008 through 2010; however, in 2014 through 2016, rural youths had 50% higher odds of smoking than did their urban peers.

Conclusions. Differential reductions in rural youth cigarette smoking have widened the rural–urban gap in current smoking rates for adolescents.

Public Health Implications. To continue gains in adolescent cigarette abstinence and reduce rural–urban disparities, prevention efforts should target rural adolescents.

Results from the 2017 Monitoring the Future annual survey show a long-term national decline in adolescent cigarette smoking.¹ Between 2000 and 2012, current cigarette smoking decreased from 14.0% to 5.0%.² Despite these declines, a 2016 national estimate found that rural adolescents had higher rates of cigarette smoking than did those in metropolitan areas.³ In state-specific studies, rural youths reported easier access to smoking tobacco products than did urban youths,⁴ initiated cigarette smoking at younger ages, and more commonly reported that they had family members who smoked and that there was smoking in their home.⁵

These rural–urban disparities matter because adolescents experience both immediate adverse health consequences with cigarette smoking (e.g., respiratory symptoms) and risk of long-term nicotine addiction, difficulty with tobacco cessation, and increased lifetime rates of smoking-related illnesses such as respiratory disorders, heart disease, and cancer. Moreover, these differences raise the question of whether rural places are benefiting from prevention initiatives at the same rate as urban areas. Between 2007 and 2014, cigarette smoking declined more rapidly among urban than among rural adults, which could reflect greater urban success in tobacco prevention, cessation, or both.⁶ However, to our knowledge, comparable analyses on the relative rate of decline for adolescents are unavailable.

METHODS

To address this gap, we studied differences in current cigarette smoking among rural and urban adolescents during 2 periods: 2008 through 2010 and 2014 through 2016. We used the US National Survey of Drug Use and Health (NSDUH) data to compare rural and urban adolescents’ current cigarette smoking prevalence in each period and to determine whether national reductions have occurred equally in rural and urban areas. The NSDUH is sponsored by the Substance Abuse and Mental Health Services Administration and provides annual data on prevalence and correlates of drug use among residents of the US civilian noninstitutionalized population aged 12 years and older. We selected study years 2008 through 2010 and 2014 through 2016 because the NSDUH changed its rural–urban identifier in 2008, and 2016 had the most current data at the time of analysis. We pooled 3 years of data at each endpoint for sufficient sample size to examine current cigarette smoking among rural adolescents. Our total sample included more than 95 600 adolescents aged 12 to 17 years, with 15% living in rural counties.

We addressed our research objectives through bivariate and multivariable regression analyses. Our dependent variable was cigarette smoking in the past month (current smoking). We used the bivariate χ² test to compare the prevalence of rural and urban adolescent current cigarette smoking in both periods. We used logistic regression to estimate the odds of current cigarette smoking, and we present results as adjusted odds ratios (AORs) with 95% confidence intervals (CIs). Our independent variables were rural versus urban residence, period, and an interaction between the 2. The NSDUH public use files classify respondents as residents of large metropolitan, small metropolitan, or nonmetropolitan counties on the basis of rural–urban continuum codes. We defined large and small metropolitan counties as urban and nonmetropolitan counties as rural.

To control for rural–urban population differences that may affect our findings, we also included gender, race/ethnicity, number of parents in the household, health insurance, and family income, because previous research has demonstrated the association of these variables with cigarette smoking. Controlling for these variables, we estimated the odds of current cigarette smoking for rural versus urban adolescents, for 2014 through 2016 versus 2008 through 2010, and the interaction between residence and period. For ease of interpretation, we used contrast statements to produce AORs for rural compared with urban youths in each period and across periods for each residence type. Per NSDUH instruction and to correct for bias associated with complex sampling methods, we used sample weights and Taylor series linearization techniques available in SUDAAN version 11 (RTI International, Research Triangle Park, NC).

RESULTS

Between 2008 and 2010 and 2014 and 2016, the unadjusted prevalence of current cigarette smoking declined for both rural and urban adolescents. However, the percentage of rural adolescents who currently smoked cigarettes was higher than that of urban adolescents in 2008 through 2010 (10.9% vs 8.3%; χ² = 30.2; P < .001), and the gap widened between 2014 and 2016 because of a smaller decline in cigarette smoking among rural adolescents over time. The rates were 7.3% among rural youths compared with 3.8% among urban youths between 2014 and 2016 (χ² = 59.8; P < .001).

The multivariable analysis found a statistically significant interaction between residence and period controlling for covariates (P < .001; data not shown). Contrast analyses indicated that, although the adjusted odds of cigarette smoking decreased for both rural and urban adolescents from 2008 through 2010 to 2014 through 2016, the magnitude of change was smaller in rural youths (Table 1). Across periods, cigarette smoking odds decreased 56% among urban adolescents (AOR = 0.44; 95% CI = 0.40, 0.48) but only 36% among their rural peers (AOR = 0.64; 95% CI = 0.55, 0.75). Reflecting this more limited rural decline, the rural–urban difference in youth cigarette smoking increased over time. Although rural and urban youths did not differ in their odds of cigarette smoking from 2008 to 2010 (AOR = 1.06; 95% CI = 0.95, 1.18), rural youths had 54% higher adjusted odds of current cigarette smoking than did urban youths between 2014 and 2016 (AOR = 1.54; 95% CI = 1.32, 1.80).

TABLE 1—

Prevalence and Pairwise Contrasts for AOR of Current Cigarette Smoking Among Those Aged 12–17 Years, by Rural and Urban Residence and by Period: United States, National Survey on Drug Use and Health, 2008–2010 and 2014–2016

	Adjusteda Smoking Prevalence, %	AORa (95% CI)
Change in cigarette smoking between periods by residence
Urban		0.44 (0.40, 0.48)
2008–2010	8.3
2014–2016	3.8
Rural		0.64 (0.55, 0.75)
2008–2010	8.7
2014–2016	5.8
Rural vs urban cigarette smoking in each period
2008–2010		1.06 (0.95, 1.18)
Rural	8.7
Urban	8.3
2014–2016		1.54 (1.32, 1.80)
Rural	5.8
Urban	3.8

Note. AOR = adjusted odds ratio; CI = confidence interval.

^aAdjusted models included gender, race/ethnicity, 1- vs 2-parent households, family income, and health insurance status.

DISCUSSION

Declines in current adolescent cigarette smoking have been smaller in rural than in urban areas, and this more limited rural reduction has widened the rural–urban gap in adolescent cigarette smoking rates. This gap persists after adjusting for socioeconomic factors associated with cigarette smoking whose distribution may differ between rural and urban adolescent populations. This result is consistent with a 2007 versus 2014 study that found that cigarette smoking declined more rapidly among urban adults than rural adults.⁶ Considering the risk of lifetime use and health consequences when cigarette smoking starts young, our findings have even greater implications for future smoking-related rural morbidity and mortality.

Youth prevention efforts may not be as widespread, or as effective, in rural versus urban places. Environmental tobacco control policies, such as tobacco-free schools and increasing tobacco product costs, are less common in rural than in urban places.⁷ Rural residents’ tobacco-related attitudes may also affect prevention efforts. For example, cigarette smoking among rural adolescents is often accepted and supported by adults^5,8 and sometimes is seen as a marker of social success.⁹ Future research should investigate the relationships between rural smoking attitudes, rural adoption of evidence-based smoking prevention programs and policies, and rural youth smoking incidence and prevalence.

These identified rural–urban differences suggest that rural places may require tailored antismoking interventions that are community led and collaborative and that reflect rural norms. For example, Vermont and Virginia have initiated Down and Dirty, an antitobacco campaign informed by rural adolescent focus groups. It aligns social marketing and event programs with the lifestyles and attitudes of a segment of rural youths by emphasizing outdoor recreation and by juxtaposing tobacco addiction with the rural values of independence and freedom.¹⁰ On the policy side, Oregon’s rural Baker County used local data and multisector partnerships to cultivate support for an ordinance prohibiting smoking in all of its town parks.¹¹

PUBLIC HEALTH IMPLICATIONS

Along with higher youth smoking rates, rural communities face more limited public health infrastructure and greater reliance on external funding.¹² Public health funding initiatives should target localized interventions that reduce rural adolescents’ smoking initiation and limit their risk of nicotine addiction. Current rural adolescent prevention campaigns focus on multimedia marketing, peer-to-peer presentations, curricula on smoking refusal, and partnerships with retailers.¹¹ These programs should be evaluated and, where effective, supported by federal and state funding to reduce the future burden of smoking-related morbidity and mortality among rural residents.